Project 1 is concerned with the behavioral definition and subtyping of dyslexia. We propose to test the ability of language tests identified as predictive from kindergarten to third grade to predict reading outcome from the sixth through the eleventh grade; in particular, to test the hypothesis that different aspects of phonological processing are related to different aspects of reading and to determine the course and direction of these relationships over time in discrepant and nondiscrepant poor readers. We also propose to determine the long term (into high school) educational outcome of poor readers versus good readers, male versus female poor readers, and discrepant versus nondiscrepant poor readers. In order to address current concerns regarding the use of IQ scores in discrepancy formulaes, we have included a measure of listening comprehension which can also be used to measure discrepancy. We will also evaluate the long term reading outcome of children screened for risk of reading disability in kindergarten (including the long term impact of two types of reading instruction on reading skills). A related issue is the role of attention deficit disorder in the above analyses. To meet these goals, we will continue to test the three samples already accessed in years 01 to 05 of our current program project. These samples include a normal group, a poor reader group, and a group originally tested in kindergarten. They will be examined with a battery of tests measuring academic achievement, phonological processing, listening comprehension, general intelligence, and rote verbal learning and memory. The presence or absence of attention deficit hyperactivity disorder will be assessed by structured interview and rating instrument information obtained from parents and teachers.